[Pw_forum] Can I get dVscf without scf-consistent procedure in the el-ph calculation?

Yingli Niu niuyingli at gmail.com
Thu Sep 7 09:22:30 CEST 2006


Dear all,
  I use espresso-3.0.
  I ran example07 step by step, calculated Gamma q-point
electron-phonon interaction.
  Part of my script was
cat > al.ph.in << EOF
phonons of Al at Gamma
 &inputph
  tr2_ph=1.0d-14,
  prefix='al',
  amass(1)=26.98,
  outdir='$TMP_DIR/',
  fildyn='al.dyn'
 /
 0.0 0.0 1.0
EOF

  That is to say, I didn't write:
    fildvscf='aldv'
due to my carelessness.

  So there was I/O preblem in el-ph calculation.
cat > al.elph.in << EOF
Al el-phonon at X
 &inputph
  tr2_ph=1.0d-14,
  prefix='al',
  fildvscf='aldv',
  amass(1)=26.98,
  outdir='$TMP_DIR/',
  fildyn='al.dyn',
  elph=.true.,
  trans=.false.,
 /
 0.0 0.0 1.0
EOF

  because it did not find the file aldv.

  I know I should rerun the whole script.

  Since the the linear response equations of phonon calculations had
been self-consistent. So I had got the wave functions and their
one-order correction in case.wfc and case.dwf files.
  Can I make use of file case.wfc case.dwf or charge density file to
get aldv file (deltaVscf) without recalculate in the self-consistent
cycles?
-- 
Yingli Niu



More information about the Pw_forum mailing list